Mr. Christoph Holtmann, German Aerospace Centre (DLR), GERMANY

In electric vehicle applications, the braking technology has changed so that a part of the kinetic energy can be fed back into the battery with the traction machine. However in the case of an emergency brake the required braking power in a car is about 8 times and in commercial vehicle applications about 30 times higher than the drive power rating. This circumstance is the reason for the continued necessity of mechanical friction brakes.

In large commercial vehicle applications, retarder technologies have a long history of reducing the wear of the mechanical brakes and the maintenance costs. The combination of the electric drive train with a retarder allows to reduce the required size of a mechanical friction brake dramatically. However, the power density of retarders, especially of eddy current retarders are small compared to mechanical friction brakes. The reason for this is the need for a heavy magnetic excitation circuit, while mechanical brakes in addition to the brake disc only need a caliper.

From an electromagnetic point of view, the power density of eddy current brakes has been increased by various measures. One of the most effective measures to increase power density from an electromagnetic viewpoint is to apply a thin layer of highly conductive material to the surface of the active eddy current material, as shown in [1]. However, when this type of eddy current brake is optimized for high speeds, the power density is limited by the thermal behavior. From a thermal point of view conventional eddy current brakes are comparable to mechanical brakes because the braking power is converted to heat in a solid disc. Also in detail, both are comparable, since the skin effect ensures that the heat arises as in a mechanical friction brake only in a thin layer on the material surface of the disc. In consequence conventional eddy current brakes can never reach the power density of mechanical friction brakes. In order to reduce the possibility of overheating, a patent [2] describes a liquid-cooled eddy current brake which can also be flooded from the rotor side with water to cool the eddy-current material. The disadvantage of this eddy current brake is that the rotor rotates in the water and the torque cannot be controlled quickly. Another possibility is to place small cooling channels near the surface where the eddy currents occur, as shown in [3], but the cooling channels near the surface weaken the primary magnetic field and the torque density decreases.

In this work, an eddy current brake with a magneto-isotropic material structure that eliminates the skin effect is shown. The eddy currents and the heat are thus distributed almost homogeneously in the material. The material structure consists of steel pins that transfer the magnetic flux from the poles through perforated aluminum sheets [4]. Coolant flows between the aluminum sheets and the number and thickness of the sheets can be selected almost freely, thereby dramatically increasing the surface area in contact with the cooling liquid.

The work focuses primarily on the concept and the design and optimization method based on electromagnetic and thermal models for the active material as well as for the excitation windings. The electromagnetic model for calculating the torque as a result of the eddy currents is validated with an error of less than 10%. Further, the results of the optimization method show that in emergency braking more than 70% of the braking energy can be converted with the eddy current brake shown here at a power density of approx. 9 kW / kg.

[1] Anwar, S., and R. C. Stevenson. ”Torque characteristics analysis for optimal design of a copper-layered eddy current brake

system.” International Journal of Automotive Technology 12.4

(2011): 497-502.

[2] Seiwald, A., Liquid cooled eddy current brake. Publication Date

2008/03/13. WO.Patent WO2008028673 A1

[3] Flach, E., Wirbelstrombremse. Publication Date 2013/05/11.

DE.Patent DE10122985 B4

[4] Holtmann, C, Elektrodynamische Bremse. Puplication Date 2017/11/16. DE.Patent DE102016108646 B4